Comparison

"Cloud" service	PeerStorage service
Where is the data?
In the "cloud", that is on servers controlled by the service publisher.	Encrypted and scattered across the devices of other users of a PeerStorage-based service.
How is durability ensured?
The service publisher keeps several replicas at all times (typically between 3 and 5).	Data is continuously rebuilt and migrates as other users' devices disconnect. The redundancy ratio is about 1.3.
How is confidentiality ensured?
It is the service publisher's responsibility. See their terms of use.	By encryption. Only the user and those to whom they give the key can read a given piece of data.
How is storage funded?
Through the publisher's business model, which must pay for servers. Creating truly "free services" is therefore impossible.	Storage comes from resources contributed by all users. There is nothing left to fund in the usual sense, and "free services" become possible.
Who can access personal data?
The user, people they choose to share with, the service provider, and anyone the provider chooses (or is forced) to share with in turn.	Only the user and people they choose to share with.
How can one run processing over all data? (aggregation, statistics)
All processing of users' data is done by the service publisher, who has access to it.	Each user can receive processing requests on their own data. They are free to run them and return the result.
How is a service funded?
By a mix of three methods: - sale or subscription, - advertising, - collection and exploitation of personal data. The large platforms rely mainly on the third.	By a mix of two methods: - sale or subscription, - advertising. Collection and exploitation of personal data cannot happen without the user's knowledge.
Who is responsible for data shared with third parties (blog, social network, messaging)?
A service publisher can in some cases be held responsible for data that its users share.	Only the user is responsible for the data they share.
How much storage does each user get?
It is set by each service publisher's own terms.	It depends, for each user, on the resources they contribute to the community.
How can one ensure that stored data is lawful?
Report the data to the service publisher, who is legally required to remove it and may have to identify the user. That is an advantage when the law is legitimate.	Nothing can guarantee that stored and exchanged data is lawful. That is an advantage when the law is illegitimate and freedom of expression is restricted — which is the case in much of the world.
How can network neutrality be guaranteed?
Only in a limited way through regulation. ISPs are the main lobby for deregulating the Internet to allow discrimination.	Traffic is technically undifferentiated, so ISPs or other intermediaries cannot threaten net neutrality.
Who owns the Internet?
The Internet becomes the property of a few firms in an oligopoly. They treat it as a way to reach their "customers" or, for Type 3 services, their "products".	Every user contributes resources and is not only a consumer. The Internet becomes everyone's property — a "commons".
Energy use?
Keeping data and keeping the Internet running requires powering data centres. Their energy use is already significant and is set to grow.	Spare capacity on each Internet user's devices forms a shared storage pool. The need for data centres — and their energy use — is reduced.
How does one find a piece of data?
Access uses memorable URLs such as http://domain.com/page, which depend on domain names and routing to a server.	Access uses a non-memorable URL such as ps://8Gg6e5ql0d, which does not depend on a domain name and does not point at a particular server.
How long is data stored?
As long as the service publisher keeps their servers online. Many things can end a service: the publisher's decision or bankruptcy, state censorship, an attack, and so on.	Independently of any third party, for as long as the user has the right. They earn those rights in proportion to the resources they contribute when using a PeerStorage-based service.
What if identifiers or passwords are lost?
A service usually provides procedures to recover an identifier and reset a password.	Only the user's key unlocks their data. They alone are responsible and must avoid losing it. Otherwise the data cannot be recovered.
Resistance to hacking?
A service publisher holds a large amount of personal data in clear form. That makes them a prime target: attacks concentrate on them because the payoff is high relative to the cost.	A user's data is scattered at random among many hosts and is unreadable to those hosts. Encryption and dissemination make the cost/benefit ratio of attacks prohibitive.